Artificial Intelligence
And I don’t mean that in a bad way…
“The AI-Powered Enterprise,” by Seth Earley, was sent to me because of my focus on artificial intelligence (AI) and machine learning (ML). While it does cover, at a high level, how AI can be of use in many areas of business, its main benefit is to help educate middle and upper-level management about the digital age. We in high tech often forget how many people are still confused, and this book serves an important purpose. Well, two purposes.
I’ve always been amused by the phrase “Big Data.” There’s an old aphorism that data expands to meet available storage. We’ve always been dealing with as much data as we can. The first few chapters are good at putting forward the importance of data. They aren’t limited to AI, as I’ve seen this discussion happen for the decades I’ve been in the industry. The change to the internet economy has meant a large expansion of data, and that means the importance of managing it now seems to be impinging seriously at the upper levels of companies.
The core of the book is a description of, and a statement about the importance of, ontologies, and the first two chapters deal with an introduction of the concept. If you go to a dictionary, you’ll find a fuzzy description on metaphysical concepts of groups. Ignore that. In data and analysis, an ontology is a categorization of high level terms that drive the organization of information. There are multiple definitions of how ontologies relate to taxonomies, categories, or more; but stick with the management concept of groups of terms that are related. For instance, the finance group will have a different ontology than does manufacturing, just as car manufacturers have a different one from logistics companies.
While I think the author should have spent more time focusing on the high level importance of ontologies, the Applied Materials example (page 35), is a nice example to put the concept in a business context.
Chapter three is good in that it puts an old, and often ignored, concept at the forefront of what is to come. People talk about “customer experience” without realizing what it means about the many touchpoints a company has with prospects and customers. The amount of data business is now collecting means that AI can be applied to better understanding the customer experience. I’ve often talked with people about the need to integrate the aspects. Figure 3-2 is nice in that it shows how the business flow and the IT needs overlap.
The rest of the chapters take a look at different business areas, and discuss both the information that should be considered and how AI can help in understanding the challenges. Chapter five has one of the few issues I’ve seen. The discussion of chat for search seems to imply a swap to almost exclusively chat. It is going to be a more gradual transition, and I see what Mr. Earley suggests being layered under existing searches and filters to get a quick boost, with chat coming later. The AI underpinnings of the search and analysis don’t require a chat front end.
My favorite chapter is six, about the sales process. Sales has always been a struggle to manage and many salespeople are why. They want independence while management wants repeatability. The chapter does a good job of showing how AI can help in both flexibility and tracking. In another basic graphic, Figure 6-1 is a simple table linking how signals from multiple system sources can be leveraged by AI to provide sales insight.
In support of the original statement that this is more of a general management book than an AI one, I point to chapter seven on customer service. The discussion surrounding Figure 7-1 is good regardless of AI.
The other problem I have with the book is chapter eight, on human resources. I have a lot more to say on it, but this isn’t the forum. What I’ll point out is his example of training a system based on resumes used in hiring (page 186). Bias has regularly been discussed in this column and elsewhere, and is a serious issue; yet it isn’t mentioned. HR is one of the last places I’d consider AI, and we need far more clarity and regulation before that should happen.
The final two chapters are good, and they do deal with helping people understand AI’s place in the organization. However, they serve as good ideas for any organization working to handle environmental change. Managers still working to adapt to ecommerce, IoT, and other technologies will find this valuable, even if they’re not putting AI at the forefront.
“The AI-Powered Enterprise” is for management interested in better understanding how to adopt new technologies and the processes needed to link line organizations with IT. The book uses artificial intelligence as a tool to describe those things, and it shows how AI can help throughout the business environment, but don’t think of this as an AI book – It’s a management book.
